Eindhoven University of Technology MASTER Characterization of the deformation behavior for light responsive liquid crystal bilayers Foelen, Yari

(1)

Eindhoven University of Technology

MASTER

Characterization of the deformation behavior for light responsive liquid crystal bilayers

Foelen, Yari

Award date:

2018

Link to publication

Disclaimer

This document contains a student thesis (bachelor's or master's), as authored by a student at Eindhoven University of Technology. Student theses are made available in the TU/e repository upon obtaining the required degree. The grade received is not published on the document as presented in the repository. The required complexity or quality of research of student theses may vary by program, and the required minimum study period may vary in duration.

General rights

Copyright and moral rights for the publications made accessible in the public portal are retained by the authors and/or other copyright owners and it is a condition of accessing publications that users recognise and abide by the legal requirements associated with these rights.

• Users may download and print one copy of any publication from the public portal for the purpose of private study or research.

• You may not further distribute the material or use it for any profit-making activity or commercial gain

(2)

(3)

(4)

MASTER THESIS

Characterisation of the deformation behavior for light responsive liquid crystal bilayers

Author Yari Foelen

Supervisor Marina Pilz Da Cunha

Graduation committee Prof. Dr. Albert P.H.J. Schenning

Dr. Michael G. Debije Prof. Dr. Nico A.J.M. Sommerdijk

Marina Pilz Da Cunha

Eindhoven, February 2018

Department of Chemical Engineering and Chemistry Stimuli-responsive Functional Materials and Devices

(5)

(6)

Abstract

Photo-responsive liquid crystal actuators are of great interest for remote switches, smart sensors and soft-robotics. To enable precise design in each application, a fundamental understanding of this deformation behavior is required.

This thesis focuses on the properties that contribute to the rate and amplitude of deformation of splay aligned liquid crystal actuators. The timescale at which actuation happens is directly correlated to the rate of heating, as such the rate can be adjusted by changing the heat generating components or the parameters such as conduction, convection and heat capacity of the liquid crystal network or the environment. The deformation amplitude as a function of incident energy or temperature displays two linear regions as a result of the total thermal expansion difference between the homeotropic and planar side of the liquid crystal film. The temperature at which the response of deformation amplitude changes is determined by the effect of photo-softening on the glass transition temperature. This allows the prediction of the deformation response by characterization of the thermal expansion coefficients of the liquid crystal film.

Lowering the crosslink density of liquid crystal films does not change the timescale of deformation but has an impact on the rate and displacement due to the increased expansion coefficients and the lower T_gcombined with an increased photo-softening effect.

Bilayer structures that combine the actuation of liquid crystal films with other static polymers can be of interest in different applications. In this work, PDMS is combined with a photo-responsive liquid crystal film in a bilayer design. Functionality can be implemented by doping of the static polymer (PDMS) with functional fillers, creating a composite layer.

This doping has minor impact on the performance of the bilayer as the modulus of a PDMS-composite stays in the same order as the modulus of the PDMS. The optimal bilayer design was investigated and the influence of a layer of PDMS on the liquid crystal film was analyzed for its effect on the rate and amplitude of deformation.

Liquid crystal films used in bilayers should have a splay alignment with the PDMS(-composite) coated on the homeotropic side of the film for maximum actuation potential. The PDMS layer affects the deformation rate by acting as a heat sink, this slows down the heating and thereby increases the time scale of deformation. The deformation amplitude of bilayers is controlled by the stiffness of the PDMS layer, as such increasing the thickness of this layer imposes an increasing restriction.

(9)

(10)

1. Introduction

Recent developments in responsive liquid crystal networks have led to many applications and opportunities for the use of their actuating properties in remote switches, smart sensors, soft robotics and other devices by adding functionality through the implementation of liquid crystal films in a layered structure.

The mechanisms that lead to mechanical deformations of liquid crystal polymer films have been extensively studied by various researchers over the last decade among which include D. Broer, T. Ikeda and T. White. Many of the contributions to those deformations have been clarified and modeled, but not much is known about the actual performance features that liquid crystal films can deliver and how to modify the deformation response.

Although liquid crystal networks have been used in multiple layered systems before, the resulting deformations in these systems were the result of trial and error or based on empirical knowledge as a fundamental understanding about these deformations was not yet quantified.

The precise design of stimuli-responsive (multi-)functional devices based on liquid crystal films in bi- or multilayers brings along the need to define the features that determine the deformation behavior for bi- or multilayers incorporating liquid crystal films. The characterization of the contributions that control the deformation through a systematic study provides fundamental insights necessary for the development of further applications.

1.1 Justification of the project

Within the Functional Organic Devices and Materials group, many photo-responsive liquid crystal systems were studied and characterized throughout the years. The contributions of both thermally and photo-generated mechanical deformations are coupled through the isomerization effect^[1]and photo-softening^[2]. In contrast to metal bilayers for which there are mechanical deformation models^[3], deformations of bilayers that contain liquid crystal films are not straightforward because the mechanical response is a combination of bending the added layers with the deformation of the liquid crystal film’s photo-thermal response. To further develop the use of liquid crystal films as actuators with added functionality by the introduction of different layers, a qualitative understanding of both the contributions to liquid crystal film deformation behavior and the influence of added layers on a liquid crystal film is required in order to modify the actuation behavior and predict the design for systems with specific performance demands.

1.2 Approach

The aim of this research project was to characterize the deformation behavior for liquid crystal films and bilayer systems, this was done through video- and thermal analysis to quantify the mechanical and thermal response. The deformation of liquid crystal films was investigated by characterization of the mechanical properties (modulus, T_gand thermal expansion coefficients) and their temperature response. The mechanical properties were correlated to the deformation analysis that coupled the deformation behavior and temperature.

Thereafter, LC-PDMS bilayers were produced to define the influence of added layers with various thickness on the deformation of the resulting bilayers.

(11)

(12)

2. Theoretical background

2.1 Liquid crystals

Liquid crystals (LCs) are a form of thermodynamically stable matter with properties between the liquid and the solid phase. Thermotropic LCs form a thermally induced mesophase when heated above melting temperature resulting in an ordered molten phase and when heated above the clearing point an isotropic liquid is formed^[4]. Different mesophase states exist in a temperature range between an isotropic liquid and a crystalline solid. The LC phases are defined by the degree of order: the nematic phase has directional molecular order but no positional order, smectic phase has directional order as well as positional order in 1 dimension and can be seen as a layered structure for which each layer is a 2D liquid^[5] (Figure 2.1). More variations of LC phases exist but are not relevant to this research project as only nematic phase LCs were studied. One liquid crystal composition can exhibit multiple phases, depending on the conditions^[6]. As thermodynamic laws predict, the phase with a higher degree of order will occur at a lower temperature.

Figure 2.1: Liquid crystal phases: isotropic, nematic, smectic.

Adapted from Liquid Crystal Devices with In-Plane Director Rotation by C. Desimpel, 2006

Liquid crystals are composed of anisotropic molecules of which most are either rod-shaped or disk-shaped. The aspect ratio of the molecules that leads to anisotropic properties induces a certain order in the mesophase under appropriate conditions. Typical LC molecular shape can be summarized as two substituted ring systems connected by a linkage group^[7](Figure 2.2).

Figure 2.2: Liquid crystal molecule simplification.

(13)

2.1.1 Liquid crystal alignment

Liquid crystals in a mesophase state have no long-range order and tend to form multi-domains but with the aid of a directing coating on glass substrates, the LC molecules are aligned over a long range. Depending on the type of coating, the LC molecules are directed parallel (planar) or perpendicular (homeotropic) to the substrate^{[8? ]}. Combining two substrates with spacers forms a cell, this leads to different alignment possibilities (Figure 2.3):

1. Planar - planar: all molecules are directed parallel between the substrates

2. Planar - planar with an angle: molecules are parallel and twisted over the profile between the substrates, this is called a twist. When this twist is 180^◦the configuration is anti-parallel.

3. Planar - homeotropic: Molecules are parallel on one substrate and gradually the director goes to perpendicular on the other substrate, this is called a splay.

4. Homeotropic - homeotropic: all molecules are directed perpendicular to the substrate.

Figure 2.3: Alignment configuration possibilities.

a) Planar b) Twisted planar c) Splay d) Homeotropic

Other alignment techniques exist such as photoalignment which uses linearly polarized light^[9]. Alignment can also be directed by applying an electric or magnetic field over the cell^[10]. 2.1.2 Liquid crystal networks

To obtain free-standing LC films, the LC monomers are linked in a densely crosslinked network by photo-polymerization. The crosslink density is controlled by the ratio of LC monomers (=

mesogens) with single functionality and with double functionality (mono- and diacrylates)^[11]. First, the LC mixture is brought to the right temperature to obtain a certain phase. Then, UV polymerization of the LC mixture fixates the order and alignment of the system by crosslinking of the mesogen molecules^[12]. This results in a glassy LC film with a relatively high modulus (1-3 GPa)^[13].

4

(14)

2.2 Liquid crystal film deformations

Liquid crystal networks (LCN) exhibit thermal anisotropic expansion perpendicular to the direction of the long axis of the molecules due to the increasing intramolecular distance to reduce order (increasing entropy) and compression along the long axis of the molecules. This leads to mechanical forces within the network that results in different deformations when the LCN molecules are not uniaxially aligned.

The splay alignment leads to bending over a single axis^[14] where both sides of the LC film contribute to the same movement. The planar aligned surface contracts and the homeotropic aligned surface expands, resulting in a smooth bending of the film (Figure 2.4).

In addition to the thermal expansion, the modulus of the LC decreases at higher temperatures, lowering of the stiffness results in higher deformation. When a splay LC film is heated below the melting temperature, reversible deformation will occur.

Figure 2.4: Splay alignment with corresponding deformation behavior.

Nematic splay LCs correspond to deformation in one plane, along the long axis of the planar direction.

Adapted from Responsive Liquid Crystal Networks by C. Van Oosten, 2009.

LCNs exhibit a photomechanical effect through the implementation of photo-active molecules in the LC network which can be switched between isomerization configurations by a specific range of wavelength^[15;16]. A popular class of molecules for adding photo-responsiveness to LC films are azo-benzene chromophores: these molecules have a ground state trans-conformation which enables easy incorporation along the LC network alignment. Photoisomerization induces the cis-conformation resulting in a reduction of the order in the LC network, causing uniaxial contraction along the molecular director and expansion in the plane perpendicular to the director^[17]. Even at small concentrations of azo dye, significant deformations can be obtained^[18](Figure 2.5).

Figure 2.5: Molecular azo-isomerization with corresponding macroscopic effect.

Adapted from Responsive Liquid Crystal Networks by C. Van Oosten, 2009

(15)

For nematic LC films with splay alignment, the expansion and contraction at each side of the film due to photoisomerization of the azobenzene chromophores is opposite and therefore constructive for the total deformation^[13], in the same way as it is for the thermal response.

Expansion occurs perpendicular and contraction occurs parallel to the molecular direction, this means that the homeotropic side expands in the plane of the LC surface and contracts perpendicular to it while the planar side contracts in the plane of the LC surface and expands perpendicular to the film surface (Figure 2.4). The light-driven in-plane bending deformation offers the possibility for the use of LC films as artificial actuators.

Figure 2.6: Deformation behavior caused by azo-dye isomerization for splay aligned LC film.

Adapted from Responsive Liquid Crystal Networks by C. Van Oosten, 2009

The absorption bands of the cis and trans form can be tuned by placing various substituents on the azo-benzene dye, these absorption bands range usually between the UV and visible spectrum^[17]. The substituents on the azo-benzene dye also influences the thermodynamic stability of the isomerization configurations, as expressed in the half-life time of different azobenzenes^[16].

The light absorption by the azobenzene chromophores causes a gradient in light intensity throughout the film, as such both the LC film thickness and the azo concentration influence the deformation as contributing factors for light absorption throughout the film^[19]. The splay alignment adds to this gradient effect as planar oriented molecules absorb more light than the homeotropic oriented molecules. Together with the difference in expansion/contraction directions between planar and homeotropic oriented molecules, this light absorption gradient results in the photoresponsive deformation of LC films^[20].

An excited azo molecule will try to retain its thermodynamic most stable state in the dark, this conversion to the trans state happens by thermal isomerization: the excessive energy in the molecule dissipates as heat. This way, the photo-induced deformation is fully reversible.

Thermal energy is delivered by the cis-trans decay of the excited isomers as well as by the environment^[21], chromophores with a short half-life time cause more heat development and a bigger light and heat gradient throughout the film as the molecules closer to the surface are fast to absorb and dissipate energy again.

Therefore, when illuminating splay LC films containing azobenzene chromophores with high-intensity lamps the mechanical response is the results of a photothermal effect. This photothermal mechanical deformation response can be used in applications as a remote controlled actuator where the deformation serves a purpose.

6

(16)

2.2.1 Photo-softening

Illumination of LC films with light of a wavelength within the absorption spectrum causes deformation at temperatures far below the T_gof the material. Several studies^[2;22]have proven that the temperature at which the storage modulus drops, typically linked to T_g, decreases and a lower rubbery plateau is reached under illumination. This shift in mechanical properties under illumination is commonly referred to as photo-fluidization or photo-softening.

Vapaavuori et al.^[23] dedicates photo-softening to the local environments created by the cis-trans modulation of the azobenzenes as this increases free volume and creates a local environment with a temperature far above T_g. These local events were studied with IR spectroscopy^[23] and proved the theory that illumination produces a stream of randomly occurring photon absorption events, each of which induces a local glass transition by releasing energy that creates a high local temperature which leads to a fluidization event that melts the local environment^[22].

The result of photo-softening measurable on a macroscopic level is the shift in storage modulus for the observed surface temperature during illumination, which is a fraction of the generated local temperatures. This shift in storage modulus is part of the shift in mechanical properties and also applies to the T_g.

As photo-softening is partly considered as a free volume effect, the crosslink density of the LC network influences the creation of free volume. This was confirmed by Kumar et al.^[2], as they showed an increased shift in mechanical properties for lower crosslinked networks under illumination.

(17)

2.3 State of the art

2.3.1 Historical overview liquid crystal actuators

One of the first studies describing a photomechanical polymer response dates back to 1967.

The photoviscosity effect, as it was named 50 years ago, describes the isomerization conversion of an azo dye to cause a change in geometry and the possibility to influence the conformation of polymers^[24]. More than a decade later, the photomechanical effect induced by the isomerization of aromatic azo chromophores in polyethyl acrylate networks was investigated by Eisenbach^[15].

In the early 90’s, the incorporation of chromophores in LC networks was studied with the intention to use the photo-chemical phase transition for data storage^[25]. Also, the anisotropic thermal expansion of LC films for controlled actuation was studied by Broer and Mol^[26]. The beginning of the 21^st century preluded a breakthrough in the field: the opto-mechanical effect in solids was described by Finkelmann et al.^[27] that explains the reversible shape changes induced by photoisomerization of nematic elastomers through an empirical and molecular analysis, followed by a model for the dynamic response that coupled the thermal and photonic effects. The two mechanisms are coupled through the isomerization hole-burning effect: when the population fraction of trans isomers is depleted, the material is less absorbing and allows light to travel deeper into the material^[21].

A wider and in-depth study on the photoinduced motions and underlying mechanisms was performed by Natansohn and Rochon^[10] and a range of photosensitive materials and the effect of their proportions in a LC system were linked to the mechanical response by Hogan et al.^[28]. In 2003, Ikeda et al.^[29] linked the asymmetric mechanical response to the light gradient in the uniaxial alignment of the network.

Further understanding of the different influences and their effect on the deformation behavior was gained over the years, correlating polarization^[9], light intensity^[30], irradiation time^[31], heat generation^[20;21], matrix composition ratios^[11], LC alignment^[32], sample dimensions^[33], different azo chromophores^[28], the concentration of chromophore^[34] and sample preparation^[29]to the mechanical deformation of LC films.

Larger bending is obtained for non-uniaxial alignments such as splay or twisted aligned LC films as shown by Broer and Mol^[26]. This type of alignment was first used as a photoresponsive actuator in 2005 by Harris et al.^[35]and since then splay and twisted aligned LC films were used in all sorts of actuator applications.

Kumar et al.^[36] described a fluorinated azobenzene doped LC film capable of continuous chaotic oscillatory motion when exposed to sunlight. This effect was only observed when blue and green light was simultaneously present, suggesting that the continuous motion results from the dynamics of the forward and backward isomerization switching.

Another oscillating LC film was studied by Gelebart et al.^[37]for which the oscillating behavior is proven to be the effect of self-shadowing in combination with a very fast isomerization half-time. The ability of continuous deformation in LC films is desirable with the prospect of devices that perform continuous motion under a constant stimulus.

8

(18)

2.3.2 Liquid crystal film deformation in applications

In 2007, van Oosten^[14]showed that the highest amplitude for the bending radius is observed for splay and twisted aligned liquid crystal films. For these films, the bending direction is dictated by the director orientation and independent of the angle of the incident light. For each alignment type, the bending radius can be precisely tuned to use in microscale devices by tuning light intensity, material composition, and film dimensions^[38].

A few years later in 2009, van Oosten et al. demonstrated a microdevice with self-organizing liquid-crystal network actuators, fabricated by inkjet printing technology. Light-driven microactuators with different subunits embedded in the LC network allows selective actuation controlled by the wavelength of the light so that the microactuators mimic the motion of natural cilia (Figure 2.7). The microactuators have the potential to create flow in a wet environment for lab-on-chip applications and the production process for these miniaturized active polymer systems can be adapted for large-scale roll-to-roll production at low cost^[39].

Figure 2.7: Artificial light-driven cilia produce an asymmetric motion controlled by the spectral composition of the light.

Adapted from Printed artificial cilia from LC network actuators by C. Van Oosten et al., 2009.

A lot of creative applications have been developed over the last few years, a peculiar example is the light-driven artificial fly trap by Wani et al.^[40]. This device works autonomously by using optical feedback to trigger the photomechanical actuation: an optical fiber shines light through a hole in the LC film and when an object enters the beam close enough to the LC film, the light is reflected and reaches the LC film which results in bending of the film (Figure 2.8).

Figure 2.8: Schematic drawing of the light-triggered artificial fly trap.

Left: At its open stage, when no object has entered its field of view, no light is back-reflected to the LCE actuator. Right: The fly trap closes when an object enters its field of view and causes optical feedback to the LCE actuator. Light-induced bending of the LCE leads to closure action, thus capturing the object Adapted from A light-driven artificial fly trap by O. Wani et al., 2017.

Another application is the four-blade mill developed by Vantomme et al.^[41], for which the actuation of LC films is converted into continuous rotation by constantly shifting the center of gravity due to the bending of the illuminate blade. The equilibrium wants to be restored by rotating but this places the next blade in the light beam, this causes a successive bending/unbending of the different blades which results in a continuous rotation (Figure 2.9).

(19)

In that study, hydrazones were used as photo-switches and were found to be an interesting alternative for azobenzenes.

Figure 2.9: Schematic representation of an LC mill application.

The mill turns counterclockwise upon light radiation on consecutive blades.

Adapted from A four-blade light-driven plastic mill based on hydrazone LC networks by G. Vantomme et al., 2017.

Recently, a different type of continuous motion with LC films is obtained by incorporating azobenzene derivatives with fast cis-to-trans thermal relaxation. These films exhibit continuous, directional waves under constant light illumination driven by self-shadowing.

The oscillating photoactive films are used to build a light-driven propagating and self-cleaning wave device^[37](Figure 2.10).

Figure 2.10: Photoactuated wave motion and self-cleaning surface mechanism.

Two example applications, demonstrating oscillatory transport of a framed film and rejection of contaminants. a) Photoactuated wave motion ejects sand from the surface of the film via a snap-through release of energy, demonstrating the mechanism for a self-cleaning surface b) Schematic representation of photoactuated locomotion along a flat substrate. c) d) The short ends of the active film, planar side up (c) or homeotropic side up (d), are fixed to a rigid frame. The direction of motion is dependent on the side that is exposed.

Adapted from Making waves in a photoactive polymer film by A.H. Gelebart et al., 2017.

10

(20)

A multilegged gripper constructed by Shahsavan et al.^[42] mimics the self-peeling of gecko toe pads by combining of gecko-inspired adhesives and LC films (Figure 2.11). The LC films with the adhesive on one end can be used to pick up objects, with the aid of magnetic patches a preload stress is induced for proper adhesion. By thermal actuation of the LC film, the adhesive parts peel off and the object is released (Figure 2.12). A remarkable load bearing capability was observed for the LC films, equivalent to carrying loads up to 100 times their own weight within their range of deformation.

Figure 2.11: Schematic view of a multilegged LCN-based gripper with adhesive coated ends.

Adapted from Thermally active LC network gripper mimicking the self-peeling of gecko toe pads by H.

Shahsavan et al., 2017.

Figure 2.12: Mechanism for gripping facilitated by an electromagnet and releasing induced by thermal deformation of the LC films.

Adapted from Thermally active LC network gripper mimicking the self-peeling of gecko toe pads by H.

Shahsavan et al., 2017.

Cheng et al.^[43] built a plastic microrobot for which actuation happens by visible light as a result of azotolane mesogens. The microrobot consisting out of a hand, a wrist and an arm was achieved by the combination of the films with different initial shapes and photodeformation modes. PE was added to the LC film with an adhesive to impose the necessary shapes. By irradiating different parts of the microrobot with visible light, the device was able to pick, lift, move and place milligram-scale objects (Figure 2.13).

Figure 2.13: Schematic view of states of the microrobot during the process of manipulating the object.

The arrows indicate the irradiated parts and the object moving distance is displayed.

Adapted from Fully plastic microrobots which manipulate objects using only visible light by F. Cheng et al., 2010.

(21)

2.4 Bilayer approach

The deformation of a photoresponsive LC film is a form of energy that can be exploited by using the conversion of photon energy to mechanical work in a bilayer. This allows the light-induced deformation to be combined with additional functionality. One way to integrate functional materials with LC films is in a LC film-composite bilayer (Figure 2.14). The use of a composite layer adhered to a LC film allows tuning the functionality by changing the amount or the kind of particles that are implemented in the composite matrix without changing the mechanical properties too much.

Another advantage is the low stiffness of polymer composites, resulting in minimal restriction of the LC film deformation. When these functional materials would be applied as a thin layer, the high modulus would influence the deformation drastically at very low layer thickness.

A possible polymer that can be used as a composite matrix in such a bilayer structure is PDMS which allows easy implementation of filler particles to tune structure properties and/or add functionality. Composites with conductive or piezoelectric particles in a PDMS matrix were recently studied by Indu Babu^[44]. One future application that came to mind within this project, is converting this mechanical work into an electric current by implementing piezoelectric materials. Piezoelectric materials generate current upon pressure variations which can be imposed by LC film-containing bilayer deformations^[45].

Figure 2.14: PDMS-composite bilayer principle.

Red LC film bends towards the light with the planar side. The PDMS composite (gray) is coated on the homeotropic side. The bending induces strain in the composite which generates a current due to incorporated piezoelectric particles.

2.4.1 PDMS

Polydimethylsiloxane (PDMS) is a rubber-like organosilicon polymer and is commonly used for its useful properties such as optical transparency, inertness, non-toxicity, and non-flammability. Specifically interesting within this research are the favorable electro-mechanical properties and the modulus that can be tuned within 0.8 - 10 MPa by the amount of curing agent^[46]. The modulus strongly depends on the molar mass between crosslink points, PDMS allows high control over the network crosslink density through the curing method or the amount of curing agent that causes end-linking reactions^[47]. Additionally, PDMS serves as an easy to process matrix for producing composite materials.

2.4.2 Composites

A composite is a material made out of two or more materials with the intent to combine multiple properties specific to the separate materials^[48]. With PDMS as a well-known composite matrix, a range of particles with interesting properties can be added to implement functionality such as high strength, conductivity or piezoelectricity. Previous work by Yu and Skov^[49], used PDMS and titanium dioxide nanoparticles to produce composites with improved dielectric and mechanical properties.

12

(22)

2.5 Mechanical aspects of bilayer deformation

For a LC film-PDMS bilayer, the LC film deformation is used to exert a force on PDMS as the PDMS layer has a resistance against deformation, called bending stiffness. The bending stiffness of a material is proportional to the storage modulus and to the thickness to the third power^[50]:

S= ^Et

3

12 (2.5.1)

This implies that the PDMS becomes more difficult to bend as the thickness of the PDMS layer increases.

Additionally, combining two materials in a bilayer with good adhesion along the interface can cause deformation due to the induced strain as a result of the difference between the expansion coefficients at the interface of the different materials^[3]. This effect could potentially add to the bilayer deformation, but the expansion coefficients of the LC film and PDMS are in the same order of magnitude and the storage moduli ratio between both materials is too high for this contribution to be a significant influence.

(23)

(24)

3. Experimental

The focus of this project is on the application side of developed technology within the SFD group, which allowed the use of materials and methods that are most common within this research group.

3.1 Materials

3.1.1 Liquid crystal film

The photoresponsive liquid crystal mixture studied in this work consists out of a nematic LC mixture composed of 56% diacrylate RM82, 41% monoacrylate RM23, 2% disperse red 1 acrylate (DR1A), a known commercial azo-derivative and 1% photoinitiator Irgacure 819 (Figure 3.1). The nematic to isotropic transition of the mixture is at 94^◦C. This system was also used by Gelebart et al.^[51] to study the oscillatory behavior of a light responsive LC film. The experience with this material and the fast cis-to-trans relaxation of the dye with the possibility of oscillation are the main reasons to use this type of LC films.

Figure 3.1: Liquid crystal system components.

RM82, RM23, DR1A and Irgacure 819.

Furthermore, in this study a lower crosslink density film was used. This was attained by adding 10 mol% of the inhibitor para-methoxy-phenol to the mixture. The different components were dissolved in 1ml dichloromethane and stirred at a hotplate at 30^◦C to evaporate the solvent and to obtain a homogeneous mixture.

(25)

3.1.2 PDMS and PDMS-composite

PDMS from Dow Corning Corporation was prepared by mixing the curing agent (methyl hydrogen siloxane) with Sylgard^R 184 elastomer, containing the vinyl-terminated base. For this study, a weight fraction of 10% is used as this provides PDMS layers with the appropriate stiffness to incorporate strength in the system as well as providing sufficient flexibility.

Previous stiffness characterization of PDMS was done by Seghir and Arscott^[46] for a variety of weight ratios and curing methods. For 1 mm thick PDMS samples, curing methods of 30 minutes and 2 hours at 60^◦C and 100^◦C were compared.

A PDMS composite was produced to study the effect on the modulus when a filler, representing functional materials, is embedded at a high percentage (40% volume). Titania-PDMS composites use Aeroxide^R TiO₂ T805 to study the effect on the PDMS composite modulus.

The primary particle size of TiO₂T805 is∼21 nm and overal density is 0.2 g/cm³.

16

(26)

3.2 Preparation Methods

3.2.1 Liquid crystal film preparation

LC films were made by UV polymerization of the photoresponsive liquid crystal mixture in a cell setup. To make cells, cleaned glass substrates were spincoated with polyimide; first 5 seconds at 1000 rpm, followed by 40 seconds at 5000 rpm. Two different polyimides were used to coat the substrates: polyimide AL1211 to induce homeotropic alignment and polyimide AL1051 to induce planar alignment. The spincoated substrates were placed on a hotplate at 100^◦C for 10 minutes to evaporate the solvent. Then the coated substrates were cured in an oven at 180^◦C for 90 minutes. After this, the planar inducing substrates were rubbed over a velvet surface to create microgrooves in the coating. Splay cells were prepared by gluing a glass plate coated with homeotropic aligning polyimide together with a glass plate coated with the rubbed planar aligning polyimide. By using glass bead spacers of 20 µm in UV-curing glue, a splay inducing cell for 20 µm thick LC films was made.

The cells are put on a hotplate at 95^◦C to bring the LC mixture in the isotropic liquid phase. The LC powder was carefully placed on the edge of cell opening, where it melted and filled up the cell due to the capillary effect. By setting the hotplate to 80^◦C, the LC mixture in the filled cell was slowly brought to the nematic phase. Then UV-polymerization was performed by using high-intensity UV light at 100% intensity (10 W/cm²) from a mercury lamp (Exfo Omnicure S2000) for 20 minutes to create crosslinks which result in the formation of a liquid crystal network. After polymerization, 5 minutes of post-curing on a hotplate at 130^◦C ensured full conversion of all the acrylate groups.

The cells were put in lukewarm water to facilitate the opening. With a sharp razor blade, the cells were opened by cutting through the glued areas. Most of the time the homeotropic side came of the glass first, which enabled cutting the LC film of the planar directing glass substrate. The obtained LC films were cut at specific dimensions (16mm x 5.3mm) for further tests.

Figure 3.2: Schematic of the liquid crystal film preparation procedure.

(27)

3.2.2 PDMS

PDMS and PDMS composites are coated with a bar coater on glass substrates, this ensures a layer with a certain thickness by using bars with a predefined gap space (Figure 3.3). The coated PDMS layer experiences a slight decrease in thickness when cured due to polymerization shrinkage. The coater bar available in the SFD lab has a coating thickness limit of 120 µm.

The coater bar was elevated by the use of paper (∼50 µm per sheet) and glass plates (∼500 µm) to obtain higher coating thickness.

Figure 3.3: Schematic of the bar coating process.

Immediately after coating the PDMS (composites) on glass substrates, the substrates were placed on a hotplate for curing. Different curing methods were used to determine which method obtained the highest modulus and the least deviation, as this is an indication of full conversion of the PDMS. Tested curing methods were 60^◦C and 100^◦C for 2 hours and room temperature for an extended time over 48 hours.

After curing, PDMS (composite) samples were cut out and removed from the glass substrate (22mm x 5.3mm) and thickness was determined.This was done using a micrometer for measuring the samples at different places to determine the uniformity of the thickness. For the thinner samples (<500µm) the height of the edges of the sample cutout was measured with the Bruker DektakXT.

3.2.3 Composite

The optimal composite has a uniform dispersity, to approach this a composite preparation procedure was adapted from Lu^[52]. To obtain a homodispersed composite of TiO₂powder in a PDMS matrix, the TiO₂powder was dissolved in isopropanol and sonicated for 15 minutes.

This solution was added to the uncured PDMS matrix and the mixture was put on the rotavap to remove isopropanol. The obtained PDMS composite could then be coated using a barcoater.

Different volume percentages of filler content (30 & 40 %volume) were produced to analyze the influence of filler quantity on the composite modulus.

3.2.4 Bilayer adherence

Good adherence between the LC film and the PDMS (composite) is required to prevent detachment of the layers due to deformation/expansion forces. When placed on a hot surface above 90^◦C, the splay LC film has the tendency to lie flat with the planar side down. This behavior made it possible to easily coat PDMS on top of the LC film (homeotropic side). With the bar coater set at 100^◦C surface temperature, the LC film was flattened on a glass substrate and PDMS was coated on top of the LC film. The glass substrate with the PDMS coated LC film was placed on a hotplate to cure for two hours at 100^◦C, this resulted in a well-adhered bilayer.

18

(28)

3.3 Characterization methods

3.3.1 Dynamic mechanical thermal analysis

Dynamic mechanical thermal analysis (DMTA) can be used to determine the mechanical properties of a material by monitoring the response when a sinusoidal deformation is applied to a sample of known geometry. The deformation is related to the stiffness and damping of the material and together with the form factor, this leads to the quantification of a range of mechanical properties of the material^[53]. The DMTA was used to determine the storage modulus of PDMS, PDMS composites and the LC films at room temperature.

For these measurements, a DMA Q800 from TA instruments was used with 0,01 N preload force. The set amplitude was 10 µm and set frequency was 1 Hz. Torque on tightening the sample clamps for LC films was 50 Nm, for the PDMS samples only 10 Nm was applied to prevent any induced sample stress from clamping. Sample dimensions were always 5.3 mm in width and length was between 4.7 - 6.6 mm. The thickness of the LC film samples was always 20µm and thickness of the PDMS samples ranged between 80µm and 1040 µm.

Also, the thermal influence on the LC film and bilayer modulus was measured as Gelebart^[54]

showed that the modulus for LC films decreases along temperature increase. This was done over a temperature range of -20^◦C to 110^◦C with the DMTA thermal sweep procedure.

3.3.2 Differential scanning calorimetry

Differential Scanning Calorimetry (DSC) measures heat capacity and enthalpy changes that represent phase transitions^{[55] [56]}. A small amount of LC mixture (< 5 mg) is weighed and sealed in an aluminum DSC pan. Two cycles of heating and cooling between 30^◦C and 130^◦C in increments of 5^◦C were applied. By tracking the heat flow of the sample, phase transitions can be determined as a constant heating/cooling rate requires more heat flow at phase transitions due to an additional energy (enthalpy) absorbance. DSC measurements were performed by a TA instruments Q1000 apparatus.

3.3.3 UV-VIS spectroscopy

UV-VIS spectroscopy^[57] was used to determine the absorption spectra for the azobenzene used in this research to determine the optimal wavelength for actuation of the LC films.

Additionally, the LC films and the bilayer absorbance was measured to exclude any additional absorbance of the polymer matrix. An absorption spectrum for plain PDMS was measured to confirm that it is fully transparent (Figure 3.4) over 320-800 nm. All samples were measured on a glass substrate with the glass substrate as the baseline. UV-Vis measurements were performed by a Perkin-Elmer UV-VIS/NIR Spectrometer Lambda 750 apparatus.

Figure 3.4: Absorbance spectrum of PDMS.

(29)

3.3.4 Actuation

The absorption spectrum of the DR1A azobenzene dye (Figure 3.5) shows a broad peak from around 380 nm to around 550 nm, this peak is a superposition of both trans-to-cis and cis-to-trans isomerization as the absorbance peaks of both transitions overlap^[58]. Irradiating the azobenzene with 455 nm light will provide the adequate energy for trans-to-cis isomerization, which is followed by fast thermal relaxation^[59].

Figure 3.5: Absorbance spectrum of DR1A isomerization.

By illumination with a high-intensity LED (Thorlabs M455L3-C2) at 455 nm (Figure 3.6) combined with a collimation adapter (Thorlabs SM2P50-A) for a homogeneous and adjustable light focus, the response of LC films was observed. A LED driver (Thorlabs DC4104) made it possible to regulate the energy of the focused light beam (25 - 650 mW/cm²applied to the LC film surface), so a correlation in actuation behavior and stimulus energy can be determined.

Figure 3.6: Emission spectrum of the 455nm LED used for actuation.

20

(30)

3.3.5 Video analysis

For a quantitative analysis of the LC film deformations, the video analysis system set up by Roel Van Raak and Dr. Jeffrey Murphy was used.

The setup (Figure 3.7) included the blue LED light of 455 nm (Thorlabs M455L3-C2) with collimation adapter to actuate the samples, every sample was placed at 9 cm distance from the collimation adapter. A red LED light of 617nm (Thorlabs LED4D100) at a constant intensity of 70 mA was used to illuminate the sample in the dark for optimal contrast, this wavelength is out of the absorption spectrum of the LC films. A LED driver (Thorlabs DC4104) made it possible to regulate the energy of the focused light beam so a correlation in actuation behavior and stimulus energy can be determined. The recording was processed by a Raspberry Pi model 3B equipped with a Raspberry Pi camera module V2.1 extended with a Fujinon HF25SA-1 camera lens for manually adjustable focus. All samples were placed in such a way that both lights irradiated the sample from the left as can be seen in figure 3.6. A glass box is placed over the sample to prevent airflow influences, also a black velvet cloth is placed behind the sample to enhance the contrast.

Figure 3.7: Video analysis setup overview.

Setup with a) blue LED light (455nm) for actuation, irradiating the sample perpendicular from the left b) red LED light (617nm) for illuminating the sample in the dark c) Raspberry Pi camera with the adjustable lens d) Sample holder covered with black velvet in a glass box

All video analysis samples had the same dimensions (16mm x 5.3mm) of 20µm LC films. First, the LC films were recorded over a stimulus ramp of blue light (455nm) in increments of 25 mA lamp intensity which later on was converted to power (mW/cm²), each time the stimulus was applied until the deformation reached equilibrium after which the light was switched off to allow full relaxation of the LC. Next, the same LC film sample is coated with PDMS (composite) and recorded again, which allows qualitative comparison for the influence of the bilayer thickness for each sample. The full video analysis procedure is explained in Appendix A.

(31)

3.3.6 Temperature analysis

A Xenics Gobi-640-GigE high-resolution thermal camera was used to monitor the film temperature, with the included Xeneth software the maximum temperature reached on the film surface can be determined by the included video analysis feature.

3.3.7 Integrating sphere measurements

To determine the total energy incident on the sample during photo-responsive actuation tests, an integrating sphere was used as a photometer to determine the intensity of the light over the visible-NIR spectrum^[60]. By integration over the spectrum, the total power (mW) that the sample receives at each intensity of the lamp (controlled by the current via the LED driver) was calculated. A representative sample area cutout on a black paper was placed in front of the sphere entrance and this area was exposed to light with the same focus as during the actuation analysis at the same distance of 9 cm.

3.3.8 Adjusted incident energy

The integrated sphere measurements determined the total amount of energy provided by the LED light beam to a LC film 5.3 mm wide. However, the LC film differs in shape when different incident energies are applied. To compare the absolute response for the equilibrium position at each intensity, a correction is applied to approach the correct incident energy over the illuminated surface. There are corrections calculated for either surface loss due to the sample being in a curled position, surface loss due to the sample being out of the beam or beam loss due to the sample being in the beam (Figure 3.8). These features were all determined by analysis of the images.

Figure 3.8: Side view of the LC film in different deformation possitions to display the considered corrections for the incident energy.

Light beam length in blue, beam loss indicated by the red arrows, surface loss indicated in yellow.

22

(32)

4. Results & discussion

This research investigated the contributions that affect the deformation of LC films and bilayers by combining characterization of the mechanical properties with a detailed study on the deformation through video analysis. The aim was to predict the ability of LC films to deform as a function of the mechanical properties and determine the impact of an added layer to optimize the deformation performance of bilayers.

4.1 Principles of photothermal actuation

4.1.1 Photo-softening

For characterization of the mechanical properties of the LC-films, DMTA measurements were performed for both illuminated and non-illuminated samples to track the modulus change as a function of temperature.

Figure 4.1 shows that the measured surface temperature under illumination displays a

±15^◦C shift in modulus as a function of temperature when compared to a non-illuminated sample. In addition, the rubbery plateau above T_g is lower. This effect is known as photo-softening[2;22;23;61]. It implies that the observed surface temperature at which the mechanical properties change during illumination is ±15^◦C lower than T_g observed through thermal analysis methods for the LC film used in this study.

Figure 4.1: Storage modulus shift observed for an LC film heated in an oven (dark) or by blue light (illuminated).

The observed modulus under illumination of splay films decreases from 1.6 GPa at room temperature to 20 MPa at 110^◦C.

(33)

4.1.2 Light absorption of the LC film

To determine the amount of light absorption of the LC films, the UV-VIS absorption spectrum for a LC film was measured. The amount of energy absorbed by the LC film can be estimated by integrating the overlap of the absorption spectrum of the LC film with the emission spectrum of the blue LED light of 455 nm used to illuminate the LC film. This is only an estimate as it disregards the scattering of the film and the angular dependence of the absorption; hence, this is more of an indication of the maximum energy potentially absorbed.

Integration of both absorption and emission spectra (Figure 4.2) indicates that maximum 60%

of the energy provided by the light is absorbed by the LC system to produce heating and/or actuation. If the deformation energy balance is characterized, this can serve to calculate the deformation energy efficiency.

Figure 4.2: Absorption spectrum of the LC system, emission spectrum of the applied blue LED light of 455 nm and the resulting energy absorption peak.

Black: LC film absorption spectrum.

Blue: Blue LED light emission spectrum.

Red: Integration of absorption and emission spectrum.

24

(34)

When comparing the spectrum of a LC-PDMS bilayer to the spectrum of the DR1A dye, it is clear that only the DR1A interacts with light (Figure 4.3).

Figure 4.3: Absorption spectra comparison of the LC system.

Comparison of the absorption spectra for DR1A dye, splay LC film, LC-PDMS bilayer and PDMS in the visible wavelength range.

4.1.3 Incident energy approximation

The most appropriate way to address the response of deformation and temperature as a function of energy for LC films and bilayers is to refer to the energy that is incident at the film, rather than the energy absorbed. This indicates the amount of energy needed to obtain a certain response which is an important feature towards applications. The energy provided by the blue LED lamp was measured as the energy delivered to a specific surface in the beam focal point. This surface was determined by the width of the LC films and the length of the beam.

To keep the analysis feasible, each rate response is expressed as the response to the energy delivered to the total surface, as a frame by frame correction for the exposed surface at each deformation point would be too extensive and exceed the scope of this study. Therefore, all rate determinations as a function of energy refer to the incident energy.

A correction for the incident energy was calculated based on the effective illuminated surface when the LC films and bilayers reach deformation equilibrium (calculation method is explained in section 3.3.8). As such, all amplitude determinations as a function of energy refer to the adjusted incident energy.

(35)

4.2 Actuation of LC films

The characterization of LC film deformation behavior is necessary as a foundation to properly compare the deformation behavior of LC film-PDMS bilayers.

Previous research^[13;14] showed that LC films with splay alignment exhibit the largest displacement when compared to other alignments, as the deformation occurs over a single axis by the constructive expansion of the homeotropic side and contraction of the planar side of the LC film when illuminated or heated. Illumination results in an actuation towards the planar side of the sample, independent of the side that is illuminated.

The amplitude and rate of deformation of LC films were quantified using video analysis which allowed comparison over a range of stimuli intensities by varying the incident energy of the light beam (Figure 4.4). For each applied intensity, the endpoint displacement over time as a measure of the film deformation is tracked with video analysis. With the data extracted from these videos, the rate of deformation can be derived by fitting the linear region of the plots, an example is shown in figure 4.4. The equilibrium position of the displacement for each incident energy represents the maximum deformation. The innovative video analysis method combined with the Xenics thermal IR camera enabled a better perception of the relations between displacement, rate, temperature and incident energy.

Figure 4.4: Deformation analysis data for an LC film.

Left: Deformation over time for a range of light intensities at 455 nm. The slope of the linear part (red) is considered as the deformation rate.

Right: Photograph displaying an overlay of all deformation positions for actuation with 260 mW/cm² adjusted incident energy.

26

(36)

4.2.1 Prebend

The splay LC films are photo-polymerized at 80^◦C; at this temperature, the network is crosslinked and the LC film is straight. After removing the films from the substrate, the cooling down from polymerization temperature to room temperature results in the splay LC films curling towards the homeotropic side as the homeotropic side contracts along the film length and the planar side expands. This results in an (almost) closed loop with the homeotropic aligned side at the inside of the bend for a LC film at room temperature (Figure 4.5). All subsequent experiments of photoactuation are performed by focusing the light perpendicular to the planar side.

Figure 4.5: LC film prebend and experimental setup.

The prebend LC film at room temperature defines the starting point used for the video analysis to determine the rate and amplitude of deformation. Little differences in prebend between LC film samples were noticed, probably as a result of deviations in alignment or polymerization temperature.

4.2.2 Deformation rate

The rate of deformation of the LC film induced by exposure to light is an important feature as it is the main contribution to the kinetic energy that can be converted to work in applications.

From the data obtained through video analysis, the rate of deformation as displacement over time can be fitted to an exponential curve. This implicates that almost all kinetic energy is generated at the start of the deformation when the absorbed light is converted into heat and deformation.

To understand the underlying reasons for the deformation rate response to different incident energies, the temperature of the films during deformation was recorded. For the temperature response in function of time at a specific incident energy, a similar response as for displacement was observed (Figure 4.6), especially during the first few seconds of illumination, the connection between temperature and displacement is evident. This remarkable correspondence between displacement and temperature of the initial linear region of the plot was observed for each applied incident energy.

(37)

Figure 4.6: Displacement and temperature correlation for 225 mW/cm²incident energy.

In the linear region (red line) the displacement is directly proportional to the temperature over time. The linear fitting to the deformation data represents the rate (in mm/s or^◦C/s).

The rate of deformation and the rate of heating both react proportional to the incident energy (Figure 4.7).

Figure 4.7: Rate response for deformation and heating over incident energy The rate response shows a linear trend for displacement rate and heating rate.

The deformation rate has been shown to be directly influenced by the heating rate of the sample. This suggests that changes in the heating process can directly influence the deformation rate. Changes to the heating process can be caused by changes in the heat generating components (azobenzene, photo-absorbers) and by tuning the parameters such as conduction, convection and heat capacity of the LC matrix or the environment.

28

(38)

4.2.3 Deformation amplitude

For soft robotics and many other applications, the total deformation amplitude as a response to a specific incident energy is an important property to control. The maximum deformation has an upper limit imposed by the experimental setup: self-shadowing of the sample can occur as a consequence of using a unidirectional focused light beam. This upper limit for deformation was not reached for the results presented.

The total deformation is monitored as the maximum end point displacement when the system has reached equilibrium. For each equilibrium, the temperature was determined. The effective incident energy at equilibrium, referred to as adjusted incident energy is determined by using the analysis images to calculate the corrections for beam and surface loss.

From figure 4.8, the observed temperatures at equilibrium are linear over the adjusted incident energy. Interestingly, the total deformation does not follow the same linearity as temperature but instead displays two distinct linear regions.

Figure 4.8: Total displacement and temperature at equilibrium as a function of adjusted incident energy.

The temperature reached at equilibrium is linear with the adjusted incident energy. The total displacement reached at equilibrium displays two distinct linear regions. The results of two different samples are displayed (black and red line).

(39)

A better way to look at this is the total displacement as a function of film surface temperature which clearly displays two linear regions (Figure 4.9). The point at which the linearity changes slope is found to be around ±70^◦C, which agrees with the observed effective temperature offset of±15^◦C from Tg(85^◦C) caused by photo-softening. This is the same offset as expressed by the modulus shift discussed in section 4.1.1.

The manifestation of two linear regions can be speculated to be the result of the difference in thermal expansion coefficients between the planar and homeotropic sides of the splay film.

Figure 4.9: Total displacement at equilibrium over temperature.

The slope ratio between the linear regions is±3

30

(40)

The work of Wie et al.^[62]studied a system with a composition of 60/40 wt% di-/monoacrylates of the same mesogens and observed two distinct regions for the expansion coefficients in function of temperature (Figure 4.10). The perpendicular expansion is linear with temperature while the contraction parallel to the molecular orientation displays two clear regions. The expansion perpendicular to a planar aligned film can be assumed to resemble the expansion parallel to a homeotropic film.

The actual deformation of a splay film can be related to the total difference in expansion coefficients (∆α= αexpansion- αcontraction) by correlating the deformation amplitude with the temperature of the film. This approach presumes that the deformation of a splay LC film is considered to be the result of the constructive effect of expansion on the homeotropic side and the contraction at the planar side.

Figure 4.10: a) Expansion coefficients as a function of temperature [Wie (2015)] b) the total difference in expansion coefficients as a function of temperature.

a) The thermal expansion coefficients are measured for parallel (blue circles) and perpendicular (red circles) directions of a planar aligned LC film.

b) The total difference in expansion coefficients (∆α) as a function of temperature: the slope changes above Tgas a result of the thermal expansion coefficient behavior for the parallel direction of the planar aligned film.

The total difference in expansion coefficients (∆α) over temperature displays an increase after T_g(Figure 4.10b), as a doubling in the slope of the∆α-temperature response. This is the same trend as observed for the displacement as a function of temperature (Figure 4.9), although the ratio between the linear regions in our system is higher as there is speculated to be an additional contribution.

According to the theory for bilayer deflection of bimetal thermostats^[3], this contribution could be caused by the difference in modulus or thickness between the homeotropic and the planar sides of the splay film. The deflection between two temperatures expressed as curvature is related to the following properties:

curvature= ¹

h.∆α.∆T. f (4.2.1)

f = (1+m)²

3(₁+m)²+ (₁+mn)(m²+ ¹ mn)

(4.2.2)

With h the total thickness, ∆α the difference in thermal expansion coefficients and ∆T the temperature difference. Factor f includes n=E₁

E₂ the ratio of the moduli and m=t₁

t₂ the ratio of the layer thickness.

The most dominant contribution in factor f is the difference in thickness. The modulus ratio was measured but did not reach any significant value. The contribution to the increased

(41)

slope ratio for displacement as a function of temperature is suggested to be the inequality in homeotropic and planar oriented mesogens in the splay film.

A splay LC film has homeotropic and planar alignment at the opposing surface of the film, but in bulk, the degree of homeotropic oriented mesogens is seldom equal to the degree of planar oriented mesogens (Figure 4.11). This induces a weighing factor for the expansion coefficients that determine the total deformation, meaning that the effective film deformation is defined as

∆α= υ1αexpansion- υ2αcontraction. This needs further investigation but if this is validated, the ratio between homeotropic and planar orientated mesogens in a splay film can be calculated as the deviation factor (υ₁/υ₂) of the deformation compared to the thermal expansion coefficients.

Figure 4.11: Alignment ratio effect on deformation.

The box displays a presumed orientation: a higher degree of planar oriented mesogens could result in an increased deformation as a function of temperature above Tg.

The thermal expansion coefficients of our system were measured to confirm the presumptions made based on the results of Wie (Appendix B). These results were chosen not to include in the discussion as the only successful measurement was the one parallel to the planar alignment.

The measurement of the thermal expansion coefficients parallel to the planar alignment proved the increase in displacement above T_g.

The correlation between the thermal expansion coefficients and the deformation displayed by splay LC films suggests that the uniaxial deformation of a splay LC film can be considered as a bilayer combination of planar and homeotropic aligned LC layers.

According to this theory, the deformation amplitude capacity can be predicted by knowing the expansion coefficients and, to some extent, the alignment behavior for LC films of different mixture. This allows for the design of actuators with specific response amplitude requirements.

32